FormalPara Key Points

In clinical practice, oral concomitant treatments are common in patients with chronic migraine treated with onabotulinumtoxinA.

Patients with oral concurrent migraine treatment showed a lower reduction in the number of monthly headache days, which may be explained by potential higher treatment resistance.

The use of oral concomitant treatments in patients with chronic migraine treated with onabotulinumtoxinA did not lead to unexpected safety concerns.

1 Introduction

Migraine is a complex and multifactorial neurological disease often requiring preventive treatment [1]. OnabotulinumtoxinA (BoNTA) is the first treatment specifically approved for the prevention of chronic migraine (CM) [2]. For this indication, BoNTA is injected in multiple pericranial muscles, at a total dosage of 155–195 U, every 12 weeks [3]. This paradigm of BoNTA injection has proven safe and effective for the prevention of CM in the PREEMPT trials [4,5,6] and its effectiveness has been confirmed in several real-world studies [7,8,9,10,11,12,13,14,15,16]. In these studies, the percentage of patients with a ≥ 50% reduction in monthly headache days varied from 47 to 69% [17]. This represents a very favorable efficacy outcome in the context of the available migraine preventive treatments [18].

Despite these positive findings, in some patients, the reduction in headache days with BoNTA alone is not satisfactory. The European Headache Federation states that combinations of preventive treatments should be considered in patients with unsatisfactory responses to treatment [19].

The local mode of action of BoNTA may favor the combination with concomitant treatments (CT). In the PREEMPT trial population, the use of CT together with BoNTA was prohibited [4,5,6]. Conversely, in clinical practice, the use of CT is very frequent [20]. Recently, BoNTA has been used in combination with monoclonal antibodies targeting the calcitonin gene-related peptide (CGRP) pathway with positive results [21,22,23,24,25,26].

Despite the feasibility and frequent use of CT in combination with BoNTA in clinical practice, to date, very few studies have focused on the impact of CT on migraine outcomes of patients treated with BoNTA. A Spanish multicenter study found that 90% of patients with CM treated with BoNTA used CT when the treatment was initiated, and 41% of them were able to withdraw all oral treatments during treatment with BoNTA [20]. However, no detail was provided on the pattern of use of CT, concerning the start of therapy, withdrawal, and dose change. Reporting details on the use of treatments in combination with BoNTA would provide useful guidance in routine clinical practice, especially in terms of drug interactions and therapeutic synergies.

We, therefore, performed an international, multicenter, real-world study to report detailed treatment patterns and to test the tolerability and effectiveness of oral CT in patients with CM receiving BoNTA treatment. Our primary objective was to describe concomitant pharmacological treatment patterns in patients treated with BoNTA. Secondary objectives included the comparison of monthly headache days (MHD) and monthly days with acute medication use (AMD) between patients with concomitant oral treatment (CT+) and those without (CT−), and stating side effects associated with the concomitant treatment by use of the Clinical Global Impressions Scale-Efficacy Index [27].

2 Methods

2.1 Study Design and Setting

This is a multicenter, observational, retrospective, cohort study conducted at the Neurology Departments of Charité Universitätsmedizin Berlin, IRCCS Mondino Foundation of Pavia, Italy, and the University of L’Aquila, Italy. We collected data from patients with CM receiving preventive treatment with BoNTA between January 2016 and March 2021. Data from the first four treatment cycles were obtained from all eligible patients.

The study was approved by the ethics committees of the three participating sites under local regulations (Berlin AE1/159/22, Pavia 0097925/21, and L’Aquila 34/2021). Informed consent was obtained from all patients enrolled in Pavia and L’Aquila centers, while the document was not required in Berlin because of the retrospective nature of the study. This study was performed in accordance with the “Strengthening the Reporting of Observational Studies in Epidemiology” (STROBE) statement for cohort studies.

2.2 Participants and Study Size

2.2.1 Study Groups

This study included two study groups. Group 1 included patients who received only BoNTA treatment without any concomitant oral treatments (CT−). Group 2 included all patients who received BoNTA and at least one concomitant oral treatment with a potential effect on migraine (CT+) during at least one treatment cycle. We considered anticonvulsants, antidepressants, and antihypertensive drugs as drug classes with a potential effect on migraine. Additionally, all drugs prescribed by a headache specialist for the indication of migraine prophylactic treatment were assumed to have a potential effect on migraine.

We stratified group 2 based on the indication of the CT into patients with at least one CT prescribed specifically for migraine (CT+M) and those with CT prescribed not for migraine (CT+O) [e.g., prescribed for hypertension or depression]. The decision to start, modify, or stop a concomitant treatment was based on the clinical judgment of the treating physicians.

2.2.2 Patient Selection

We screened the electronic charts of all patients with migraine who received BoNTA in the selected period. Inclusion criteria for this analysis were: (1) age ≥18 years at treatment initiation, (2) diagnosis of CM according to the International Classification of Headache Disorders, Third Edition criteria [28], and (3) health records and headache diary data availability for at least four subsequent treatment cycles with BoNTA according to the PREEMPT injection paradigm (155–195 U in 31–39 injection sites every 12 weeks) [29]. We excluded patients with insufficient health and/or headache documentation or who had already received BoNTA treatment before the observation period. Because of the retrospective observational nature of the study, we did not calculate a sample size but simply enrolled all eligible patients over the period indicated above.

2.3 Data Sources

All data were obtained from the electronic patient records. We collected general health history and headache-specific history. Currently used medications were classified into with/without a potential migraine preventive effect and prescribed/not prescribed for migraine as mentioned above. Headache-related variables were extracted from electronic or paper headache diaries. A headache day was defined as a day on which at least 30 minutes of headache was reported with a visual analog scale score (range 0–10) of at least 1. An acute medication day was defined as any headache day on which acute medication was taken. Acute medication included nonsteroidal anti-inflammatory drugs, analgesics, and/or triptans.

2.4 Variables

We evaluated demographic variables, including age, sex, migraine disease duration, the occurrence of migraine aura, prior migraine prophylactic treatments before BoNTA treatment, and current medications. From the headache diaries, we extracted the variables: MHD and AMD. For this analysis, a month was defined as a 28-day period. A MHD was defined as a day with a reported headache of at least 30 minutes and an AMD as a day with a headache on which acute medication (non-steroidal anti-inflammatory drug and/or triptan) was used. The baseline MHD and AMD values were calculated from the 4 weeks before BoNTA initiation. The mean MHD and AMD were also calculated for every 12 weeks after each of the four BoNTA treatments. Based on these variables, we calculated the 30% and 50% responder rates, which represent the percentage of patients with a reduction in MHD by at least 30% or 50% from baseline during the 12-week periods after each BoNTA cycle. Based on the electronic charts of the patient’s treatment, we assessed the side effects associated with the CT+M and scored them according to the Clinical Global Impressions Scale-Efficacy Index [27]. The side effects were ranked as “None,” “Do not significantly interfere with patient functioning,” “Significantly interfere with patient functioning,” or “Outweigh the therapeutic effect,” based on the available documentation.

2.5 Statistical Analyses

We reported continuous variables as means and standard deviations or median and 95% confidence interval (CI) for the median. Categorical variables were reported as number (%). In the case of missing headache documentation, we used the last observation carried forward approach. Patients for whom it was not possible to calculate MHD and AMD at baseline and/or from at least two out of the four follow-ups were excluded.

As our data were not normally distributed, as assessed with the Kolmogorov–Smirnov test, we used non-parametric statistics. We used Friedman’s two-way analysis of variance by rank test for repeated measurements with a Dunn’s pairwise post-hoc-test. P-values were adjusted by the Bonferroni correction for multiple testing. To assess the effect size, we estimated Kendall’s Concordance Coefficient W, where 0 (indicates no relationship) and 1 (indicates a perfect relationship). To compare groups, we applied the Independent-Samples Mann–Whitney U Test or the Independent-Samples Kruskal–Wallis Test as appropriate. A value of p < 0.05 was considered statistically significant. Statistical analyses were performed with IBM SPSS Statistics, version 28.0.1.0 (IBM, Armonk, NY, USA).

3 Results

3.1 Participants

At all three centers, 314 patients were treated with at least four cycles of BoNTA between January 2016 and March 2021. Among those, we identified 178 patients who were eligible for our study. Patient disposition and reasons for exclusion are displayed in the flow chart (Fig. 1).

Fig. 1
figure 1

Flowchart of patient selection and inclusion. BoNTA onabotulinumtoxinA, CT− patients without concomitant treatment, CT+ patients with at least one concomitant treatment

Full size image

3.2 Demographic Characteristics

Of the 178 included patients, 70 (39.3%) patients only received a BoNTA treatment without any concomitant oral treatments (CT−). The remaining 108 (60.7%) patients received at least one concomitant oral treatment (CT+).

Patients in the CT− group were on average younger compared with patients in the CT+ group (p = 0.005). On average, patients had already not responded to 2.6 ± 1.5 prior preventive treatments, which were discontinued because of a lack of efficacy or limiting side effects.

Almost half (n = 87, 48.9%) of our patients were diagnosed with medication overuse headache. Patients with CT+ and CT− had a comparable incidence of medication overuse headache (49.1% vs 48.9%, respectively). Patients with CT+, however, were more likely to experience a medication overuse headache previously (23.1% vs 5.7%, respectively). A mood disorder was diagnosed in 73 patients (41.0%) and an anxiety disorder in 61 (34.3%). Mood or anxiety disorders and hypertension were diagnosed more often in patients with CT+ compared with patients with CT− (respectively, p = 0.003, p < 0.001, and p = 0.003). All patients’ characteristics are shown in Table 1.

Table 1 Patient characteristics for study groups and strata (n = 178)
Full size table

3.3 Concomitant Oral Treatment

3.3.1 Concomitant Migraine Prophylactic Treatment

Our cohort included 74 (41.6%) patients who received at least one concomitant oral treatment prescribed specifically for migraine (CT+M) during the first four treatments with BoNTA. These patients received one (n = 53, 71.6%), two (n = 12, 16.2%), or three (n = 9, 12.2%) CT+M. Table 2 provides an overview of the prescribed CT+M drugs.

Table 2 Overview of prescribed CT+M treatments (n = 74)
Full size table

From the patients with one CT (n = 53, 71.6%), the CT remained unchanged in 31 patients (58%) during the first four treatment cycles with BoNTA. Eight patients (15%) started CT and 13 (25%) stopped CT. One patient (2%) started and stopped CT over the course of 1 year with BoNTA.

Of the patients with two CT (n = 12, 16.2%), five patients (52%) did not change the CT during the first four cycles with BoNTA. Two patients (17%) stopped one CT, one patient (8%) started one CT, and one patient (8%) started and stopped one CT. Two patients (17%) started and stopped both CTs, and one patient stopped both CTs.

From the patients with three CTs (n = 9, 12.2%), all CTs remained unchanged in two patients (22%). Three patients (33%) started a third CT and one (11%) stopped a third CT. Three patients (33%) started a second and third CT.

Side effects caused by the CT+M during the first four treatment cycles with BoNTA were reported by 15 (20.3%) patients. Only in three (4.1%) of these patients did side effects (all in topiramate 200-mg daily users) significantly interfere with the patient’s functioning according to the Clinical Global Impression-Efficacy Index. Table 1 in the Electronic Supplementary Material (ESM) shows the number of side effects and the corresponding drug per drug class.

3.3.2 Concomitant Oral Medication Not For Headache

34 patients (19.1%) receive a concomitant oral treatment prescribed for an indication other than headache (CT+O). The majority had one additional treatment (n = 21, 61.8%). Thirteen (38.2%) patients had two additional treatments. The CT+O most frequently prescribed were antidepressants in 23 (67.6%) patients and antihypertensive drugs in 16 (47.1%) patients. In the majority of these patients (n = 31, 91.2%), the treatment remained unchanged during the first four cycles with BoNTA. A detailed overview of prescribed treatments not for headache is shown in Table 3.

Table 3 Overview of treatments prescribed not for headache (n = 108)
Full size table

3.4 Monthly Headache Days

Table 4 shows MHD at baseline and their change after the first four treatment cycles with BoNTA for patients with CT−, CT+, and the strata CT+M and CT+O. In the CT− group (n = 70, 39.3%), the median MHD decreased from 20 (95% CI 19, 25) at baseline to 8.5 (95% CI 7, 10) in the 12 weeks after the fourth treatment cycle with BoNTA (p < 0.001). Patients with CT+ (n = 108, 60.7%) also showed a significant reduction in MHD from 22 (95% CI 20, 25) at baseline to 11 (95% CI 9, 14) during the 12 weeks after the fourth cycle with BoNTA (p < 0.001). Figure 2 illustrates the ≥ 30% and ≥ 50% responder rates in the total population and the CT− and CT+ groups. A reduction of at least 30% of MHD was achieved by 49 (70.0%) and 59 (54.6%) of patients with CT− and CT+, respectively, after cycle 2 compared with baseline (p = 0.043). After cycle 4, a reduction of at least 30% of MHD compared with baseline was achieved by 52 (74.3%) and 63 (58.3%) of patients with CT− and CT+, respectively (p = 0.037). We did not detect significant differences between patients with CT− and CT+ for the ≥ 50% responder rate.

Table 4 Absolute change in monthly headache days during treatment cycles 1–4 compared to baseline
Full size table
Fig. 2
figure 2

The 30% and 50% reduction of monthly headache days from baseline for patients with CT− and CT+ during treatment cycles 1 to 4

Full size image

In the stratification for CT+M (n = 74) versus CT+O (n = 34), both groups had a significant reduction in MHD from baseline to the first four treatment cycles with BoNTA (p < 0.001 for both groups). However, patients with a CT+M showed a significantly lower reduction in MHD after cycles 2–4, compared to patients with CT− (p = 0.002, p = 0.001, and p = 0.004, respectively) and to patients with CT+O (p = 0.008, p < 0.001, and p = 0.006 respectively). Figure 3 illustrates the ≥ 30% and ≥ 50% responder rates in the two population strata. After cycle 3, 26 (76.5%) and 24 (70.6%) patients with CT+O achieved, respectively, a 30% and 50% reduction of MHD compared with baseline. For patients with CT+M, these were respectively 37 (50.0%, p = 0.012) and 27 (36.5%, p = 0.002) patients. A 30% and 50% reduction of MHD compared with baseline was achieved by 26 (67.5%) and 21 (61.8%) patients with CT+O compared with 37 (50.0%) and 28 (37.8%) patients with CT+M, respectively (p = 0.012 and p = 0.024).

Fig. 3
figure 3

The 30% and 50% reduction of monthly headache days from baseline for patients with CT−, CT+M, and CT+O during treatment cycles 1 to 4

Full size image

3.5 Monthly Acute Medication Days

Table 5 shows the change in AMD from baseline and after the first four treatment cycles with BoNTA for patients with CT−, CT+, and the strata CT+M and CT+O. In the total cohort, AMD decreased from a median of 15 (95% CI 15, 17) during baseline to 7 (95% CI 6, 8) after the fourth treatment cycle with BoNTA (p < 0.001). A difference in the reduction of AMD was observed after cycle 2 between CT− and CT+M (p = 0.014) and between CT+M and CT+O (p = 0.019). During cycle 3, a difference between patients with CT+M and CT+O (p = 0.030) was observed.

Table 5 Absolute change in monthly acute medication days during treatment cycles 1–4 compared to baseline
Full size table

4 Discussion

In this multicenter real-life study, BoNTA treatment was well tolerated, and no important side effects were registered even with an additional concomitant oral preventive drug. Our data confirmed the well-established efficacy of BoNTA with a median reduction of 8 MHD after the fourth treatment cycle compared with baseline, regardless of the use of a concomitant oral preventive treatment, in line with previous reports [6, 30]. After stratification, we observed a greater reduction in MHD in patients without concomitant oral treatment after cycles 2–4 compared with patients with oral concomitant oral treatments. This observation could probably be explained by a higher disease burden in patients with oral concomitant therapies as suggested by the higher number of prior preventive treatments, and psychological and non-psychological comorbidities. These clinical features are well-known negative predictors of response to migraine prevention therapies [31].

Our cohort highlights the high frequency of polypharmacological treatment patterns in patients with CM. The majority of patients treated with BoNTA (60.7%) were taking concomitant medications. Among these, for 74 (41.6%) patients, these drugs were prescribed with an exclusive indication for migraine, while the remaining had other comorbidities for which they required medications such as anxiety or depression. Polytherapy was well tolerated, and side effects were registered in 15 subjects only (20.3%). That is significantly lower than reported in the literature regarding oral medications, which are well known for lacking persistence [32]. Most patients in our study were already taking a well-tolerated, but not sufficiently effective CT at a stable dose before starting BoNTA treatment, thus minimizing possible collateral effects. However, in some patients, the possible use of a lower-than-average dose of oral treatment (e.g., a low dose of amitriptyline as shown in Table 2, compared to the minimum/standard dose recognized as effective [19]) could lead to a lower-than-expected incidence of adverse events and a higher persistence among the patients.

Monotherapy is recommended for migraine preventive treatment based on the evidence available in the literature and clinical practice [33]. However, patients with CM are notably difficult to treat and often resistant or refractory to treatments [19]. Therefore, they often require polytherapy. For example, up to 63.3% of cases of patients with medication overuse headache were receiving polytherapy when admitted for a detoxification program in a tertiary headache center in Italy [34]. In addition, polytherapy could be justified even in patients who have a good response to migraine treatments, but a high number of residual MHD [35]. However, to date, only topiramate, anti-CGRP(-receptor) antibodies, and BoNTA are the treatments specifically approved for CM prophylaxis and few data regarding the use of concomitant therapies are available from real-life studies (Table 2 in the ESM). In most of these studies, patients receiving BoNTA were also taking concomitant oral medications, but neither efficacy nor safety was assessed by stratifying the population according to this parameter. The COMPEL study found that patients receiving concomitant preventive treatment at baseline had a significantly smaller reduction in MHD from baseline at week 108 in comparison with patients not receiving preventive treatment at baseline [36]. Casucci et al. [37] suggested that add-on therapy should be considered in patients who present with high disability [34], concurrent risk factors or comorbidities, and a history of failed preventive treatments. A recent Italian consensus paper suggests adding concomitant prophylaxis when the ≥ 50% reduction in MHD is not achieved and no improvements in any of the efficacy/benefit indicators are observed [38]. In this framework, our data support the possibility of using concomitant oral prophylaxis in those difficult-to-treat patients or those lacking a sufficient response to BoNTA, with no fear of an increase in adverse event incidence.

In this real-life multicenter study, it is also possible to observe the debatable choice of using off-label drugs that are not recommended by current guidelines specifically for migraine prophylaxis, such as gabapentin among the anticonvulsants or quetiapine, a second-generation antipsychotic. This choice probably derives from a few studies that assessed the effectiveness of the above-mentioned molecules even though there is insufficient evidence to support their inclusion in the list of effective therapies [39, 40]. However, this phenomenon points to the necessity for physicians to find acceptable alternatives to help resistant/refractory patients.

Finally, it would be interesting to address the reasons why many drugs were kept once BoNTA was added to the prophylaxis. Lacking a demonstration of the BoNTA additive effect over the other classes of treatment, this observation could reflect the fear of both patients and physicians that the improvement obtained with BoNTA may be jeopardized by the interruption.

Overall, this study is not without limitations. First, it was a retrospective observational study, although the strength of the data is supported by the use of documented health records and by the prospectively compiled headache diaries. Second, we collected data over a 4-year observation period that coincided with the advent of the CGRP-targeting monoclonal antibody. This event is likely to have introduced dynamic changes in the routine management of subjects. Moreover, the lack of consensus on the minimum/target dosage for oral migraine preventive treatments and on the duration of exposure for the assessment of efficacy may have introduced a huge variability among the different dosing schemes [19], especially when it comes to the use of these molecules for two different purposes (e.g., hypertension and migraine or mood disorder and migraine). As a real-life study, it was decided to include all the possible dosages used to depict the different attitudes toward combination therapy. To create a homogeneous dataset that was as complete as possible, we excluded patients with incomplete data or early treatment termination. Furthermore, we excluded patients with prior experience with BoNTA, as complete information about the effectiveness was not available at our centers and we preferred including “naïve” subjects only. Therefore, only those who started the treatment during the timeframe mentioned were considered for a more comprehensive analysis. This could have led to selection bias, excluding patients with poor response or with compliance issues with BoNTA and/or the other migraine preventive treatments. Moreover, the duration of our baseline (28 days before the first treatment cycle) might be too short for an accurate representative baseline, as the MHD might vary from month to month. Therefore, part of our results may have been caused by the natural course and monthly variance of the headache frequency. However, our sample size is large enough to reduce this bias to a minimum. Finally, we did not consider if CT was started before or after BoNTA treatment in our analysis. Moreover, we did not stratify for this feature in the subsequent comparative analysis. This may introduce bias, which could lead to an underestimation of the true association and the possible synergic effect of the two types of medication.

5 Conclusions

This multicenter retrospective study described the real-life clinical practice regarding the use of combinations of oral preventive medications and BoNTA in a large cohort of patients with a follow-up of at least 12 months. In our cohort of patients treated with BoNTA, concomitant migraine prevention demonstrated excellent tolerability, and we did not observe unexpected safety concerns. Patients treated with a combination of prophylaxis treatment reported a smaller reduction in MHD, possibly because of a higher burden of disease, while we did not detect any signal pointing toward safety concerns when combining oral preventive treatments with BoNTA. Further studies are needed to address the real extent of the additive effect that the combination of different drugs might have in specifically selected populations such as those with resistant or refractory migraine or medication overuse.